***The Official MSI GT80S Titan (w/desktop 980 GPU's) Owner's Lounge***

Oh... well looking at the 3DM FireStrike which puts emphasis on the GPU, 23.3K. 980Ms are not going to be reaching such a score. Maybe the Clevo 570/870 with dual PSU and 1.5Ghz+ GPU cores.

If they are referring to the GPU score in 3DM11, it's known that lower CPU Mhz/scores heavily limit the GPU scores.

My Area-51 reaches 45K+ in GPU score Firestrike and can go up to 57K for 3DM11.

 

If any other owner's would like to post their benchmarks, post similarly as I've done in post #2, and I'll place your name and link your results in the #1 post of the thread. This will provide easy reference, and insure your results do not become lost as the thread progresses.



EDIT: Also, please ensure temps are monitored, provide the system's state (stock, degree of OC, or bios mods, etc.), and the ambient room temp.

 

Striker, what does the MSI Dragon app say about power consumption? Can you run a windowed benchmark in 720P and see how much power is consumed?

 

Good question! I know a lot are curious about power and throttling. I'll see if I can get to that in a little while.

 

1.5ghz? nah not at all. This is not the highest score for 980m here, and I am already at 22.8k GPU score But yeah I wouldn't game with those clocks without a second PSU just to be safe

The 980s will shine at higher resolutions though.

 

My GT80 SLI-263 980m SLI, OC'd gets a better score than the 980 SLI score you are showing, and it's running stock BIOS/vbios - no Prema vbios:

Score 15911
Graphics Score 21345
Physics Score 13681
Combined Score 5972
http://www.3dmark.com/fs/5708173

The PM870 SLI 980m is just ahead, but still not as fast as the fastest GT80 SI-263 score posted by @Thrawn

Score 17088
Graphics Score 23592
Physics Score 13634
Combined Score 6359
http://www.3dmark.com/3dm/7834303?
http://forum.notebookreview.com/thr...es-test-and-tune.773245/page-27#post-10047597

So, what can be done to improve the 980 SLI GPU / CPU score?

@amudar could you please post the URL links to the scores on Futuremark?

 

Just wait that the 980 sli is release also for p870dm.. And then bye bye msi! Xd

 

It will likely be a frame other than the P870DM, as it has already been made clear that the current P870DM can't / won't get a 980 SLI configuration... but one can hope

"Clevo didn't changed their minds, SLI GTX980 as never planned for P870DM."
"As far as I know it's not possible, what I understand is because of the GTX980 PCB."
http://forum.notebookreview.com/thr...enix-has-arisen.781814/page-174#post-10154177

"Confirmed P870DM won't have SLI GTX980."
http://forum.notebookreview.com/thr...enix-has-arisen.781814/page-172#post-10153982

 

Ummmm..... now see here we go again..... where does Amudar indicate his GT80S is Oc'ed? That is stock!..... and of course your post brings a troll along touting Sager/Clevo. Geez guys, this is NOT what this Owner's Lounge is all about! As far as what the GT80S is capable of, who knows, because no one has attempted to find it's limits yet.



@Zero989,

I ran your request w/ 3DMark11, and throughout the test, wattage remained mostly below 140 watts..... w/ three 1-2 second peaks at 248 watts. Thus far, I have not seen any indications of throttling, and I am still yet to do any game testing yet.

 

It's all part of the process

The PM870DM has been going through the same thing...first with single 980, now with 980m SLI. It's still not there yet, and it's been a couple of weeks.

It's just a matter of pointing out the obvious and encouraging efforts for improvement.

It may be days, weeks, or months before the GT80S SLI 980 has reached it's best performance.

My GT80 SLI-263 was good out of the box, but now with the latest BIOS, several months later, I can run stable at 4.2ghz, where 4.0ghz was the best it could do out of the box.

And, next comes a SVET vbios, and hopefully BIOS down the road.

And, now the GT80S SLI 980 starts on it's path to best performance, and we are all patiently waiting

I hope you aren't saying that we shouldn't encourage GT80S performance improvements?

We all want to purchase a new laptop with better performance than what we have now, it's a matter of figuring out which one to get.

 

Whoa, 248? Something isn't right here! Thanks!

 

@hmscott, I fully understand 'the process', and since we all understand, "It may be days, weeks, or months before the GT80S SLI 980 has reached it's best performance", how is it worthwhile comparing OC'ed 980M performance? I get the sense I entirely wasted my time posting HERE, or was the point just overlooked?



@Zero989,

Yeah, I am with you, but I am sure MSI worked hard to find some semblance of balance. I know it was assumed by all that a 330W PA would be inadequate, myself included.

 

5960X @ 4Ghz with ONLY 4 cores. 20MB cache for 4 cores... mmm.

2-way SLI 980 STOCK!

nVIDIA control panel Text Filtering - Quality : High Performance <-- important

http://www.3dmark.com/3dm/9825664?

I will now match the MSI 980 SLI clocks and post that.

 

It's worthwhile comparing what we have to what we want to acquire, that's pretty basic.

It's a worthy goal to attain then surpass the SLI 980m OC scores.

Once the bar has been set, there isn't any need to keep posting it.

And, no need to discuss it any further; except to celebrate when it is surpassed

 

hmscott, here is the link:

http://www.3dmark.com/compare/fs/6678054/fs/6864136/fs/6690247#

 

Nice, another bar set to surpass

Score 17088
Graphics Score 23592
Physics Score 13634
Combined Score 6359
http://www.3dmark.com/3dm/7834303?

 

My 980s @ 1126Mhz (Boosted) I subtracted 140Mhz from my core.

http://www.3dmark.com/3dm/9825832?

I'm ahead of you by 7fps in the first benchmark, and 1fps in the second. You crush me in the physics. We're tied in the combined. LOL!

Are those 980Ms you have boosted to 1355Mhz in that run? The Broadwell L4 cache is steamrolling that CPU benchmark. You're making up ground in the combined test. Which is why these SLI 980 GT80S' are struggling. The 6920HQ is just not getting juice, and it lacks L4. :[

 

LOL Don't leave my old GT80 4980 out (over clocked to the end of 1 PSU)
16900 with 24132 graphics = http://www.3dmark.com/fs/5422608

STOCK p870dm-g score 14494 for way less than 3000 usd = http://www.3dmark.com/fs/6864275

Regardless no matter how its sliced GPU scores are nuts and now there are two manufactures producing gaming laptops with seller performance yet both machines are different in several ways. In other words don't worry be happy while this time lasts! And buy something!

 

Well here is another run. This time with full 8 cores enabled @ 4.5Ghz.

GTX 980s @ -140Mhz core to get 1126Mhz boost

http://www.3dmark.com/3dm/9826009?

Impressive memory overclock.



170 TMU?

 

Did the windows updates and most updated drivers and ran it again:

 

There ya go!

 

Question. So all things 980 SLI are based off of 1 330W psu. Is this correct?

 

Well, this is the GT80S Owner's Lounge, and it has a 330w PA.

 

@Zero989 @amudar great jobs

I noticed the CPU-Z showing a downclocked CPU - are you both remembering to run Power Plan High Performance with CPU performance set to 100%/100%?

How is the CPU OC'ing? Does the 6920HQ show up in XTU for tuning?

Keep up the good work, it just takes time and patience to keep edging those scores up

 

I think all of the benchmarks thus far reflect the same, since the default power plan is set to 'Balanced'. Mine is the same, with mode set to 'Sport'. And YES, the 6920HQ shows up in XTU..... MSI was nice to provide the XTU with the software package.

 

I don't know that's why im asking. And I know exactly what lounge this is.

 

Remember everyone, this is the owners thread for the specific GT80 with 980s desktop. There is a reason why we have a separate thread for the GT80 benchmark and performance analysis.

Keep the content here relevant for owners and general knowledge of the laptop. This is not the thread about which laptop is better, who has the highest OC score, etc.

Feel free to use the performance thread or even create another one but lets stop the comparisons ok? Too many posts already on the owners thread and most have been done by not owners at all.

Dont let me get my hammer ready, im trying to have a family dinner here

Sent from my SM-G925I using Tapatalk

 

Indeed. I am impressed with the power regulation so far.

I am also interested with the dedicated audio chip it has. Definitely a good move there.

Sent from my SM-G925I using Tapatalk

 

Not entirely sure yet, but it appears primary audio is still Realtek, and the secondary ESS Sabre HiFi is dedicated to the new gold-plated headphone jack.

 

Thanks.

 

I listened to some music earlier over the Sabre Hifi DAC with my Sennheiser headphones. I was pleasantly surprised at how good the quality of the dac actually is. I used to use a portable jds labs c5d dac and the one on this laptop provides superior sound quality over that.

 

I'm going to go ahead and assume you didn't check out the sound settings before you played music over it

You can set the quality to go higher in the sound settings... MUCH higher than default

 

Ugh..... as far as the Sabre HiFi, my Steelseries headphone are wireless, so no joy..... but, I bet they'll sound great anyhow. Time to jump on Steam, and get busy with some gaming.

 

So about the Saber Hifi Audio DAC.

1. Excellent Sound quality and it can be increased in the Sound Options of your computer as it is its own device.
2. It does not work well with 3.5mm jacks with three rings. I would go as far to say only use a jack that supports the left and right audio channels
3. It does not support voice channels ( Need to do more testing, but I don't think it does so far)
4. Sounds great for games so far!

 

For those of you that don't know where that thread is, it's not been posted to since November 17th, please find it here:

MSI GT80 3dMark Scores, test and tune
http://forum.notebookreview.com/threads/msi-gt80-3dmark-scores-test-and-tune.773245/

It might be time to rename that thread, or start a new one for all GT80(S)'s?

We are all anxious to get benchmarks and tuning info for all GT80(S)'s, as we are all in this together on the path to upgrading, current and future owners alike

 

It might be a bit too fine of a point to run Stock on Balanced - at least AFAIK everyone runs at High Performance - the stock term means the best that can be done using stock BIOS/vbios and no OC settings.

High Performance hasn't usually been considered an OC setting. And, it does make a real difference in getting the best results at Stock testing.

Skylake under Windows 10 and the new Step Shift might be removing this distinction, at least it is supposed to make Balanced perform as well as High Performance.

It would be fun to find out if there is still a difference between Balanced and High Performance on Skylake Windows 10 with Speed Shift.

Would you please post a screen shot of stock XTU 6920 Tuning page? It would be nice to see stock values and what can be adjusted.

 

Sorry, but now you are going to tell me what 'stock' means? Once again, I always do initial benchmarks of what users can expect stock, 'out-of-the-box'..... and then later, run benchmarks at the 'High Performance' setting. I'll run these when I have time. Right now I have games installing from Steam. As to XTU screenshots, you can find these in post #2.

 

It's noted that for this thread, and your benchmarks in post #2, stock benchmarks use the Balanced Power plan.

Thanks for adding the updates to post #2. I will look for updates there as well as this end of the thread from now on.

I look forward to the benchmarks run on Stock settings under High Performance power plan for comparison.

Thanks -=$tR|k3r=-

 

Just for you, and as expected, the 'High Performance' power plan results in only a marginal improvement.... 76 points.

 

Cool, thanks.

This means that Skylake Speed Shift is indeed helping Balanced to react quicker to CPU demands and giving almost the same results as High Performance.

That also might be much of the difference between @amudar first and 2nd run posted scores, after he fully updated and got the Skylake Speed Shift update applied to Windows 10.

Thanks again

 

With my last GT80 SLI-009, MSI's Speed Shift was set to 'Comfort Mode', by default, out of the box..... so changing to 'Sport Mode' (a mild OC) made a difference, but then as now, changing to the High Performance power plan not so much. With the new GT80S, MSI has Speed Shift set to 'Sport Mode' by default, out of the box, and once again, the High Performance power plan changes little.

 

Ahhh, ok, we are really coming from two different frames of reference when it comes to "stock" vs "OC", due to the MSI Dragon Gaming app - you use it - and I don't.

 

Now see, this is as far as I go. What is it you really seek here? There has never been a debate of what 'stock' is, or 'stock vs. OC'..... only you trying to make an argument or a point for such. Since it seems to have zip-lined over your head, I'll write this one last time. My initial benchmarks are done 'stock', no optimizations other than software updates, with default settings and MSI's installed software. I do this to best represent what user can expect of the 'out-of-the-box' experience. As to the Dragon Gaming Center, this is a part of MSI's software installation, is it not? Why is this a difficult concept for you, and why continue with this rhetoric? Since you so desire to espouse yourself, why not become an actual owner of this notebook, and then become truly helpful?

 

Nevermind

 

Something I would like to confirm from the owners if possible, are the CPU multipliers locked? If not how many multipliers are open for each core in XTU? And last if open multipliers are present will it hold those clocks? Thanks in advance!

 

I have a Micro answer. I noticed that in CPU-Z my processor was running way above stock. Something like floating around 3.5ghz+

 

@-=$tR|k3r=-

Hi


Remember that benchmark I posted using the 850 Pro 2TB... ?

Well, I did more reading and the results were... INSANE.

Read/Wrote speeds jumped 10x times what they were and all I did was activate "RAPID" for that drive. Went from 540 Read/ 490+ write to...

Jesus...

5k read/ 4.8k+ write...o__o I will repost
results.

 

Eye candy - you mostly test your RAM speed working as cache.

 

Can't argue with the results... NOPE!

 

Good numbers, too bad rapid mode doesnt help much in real world performance. As zipperi said, its mostly RAM cache